Christian Batoon

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Top Label Types

Audio Recording

Bounding Box

Classification

Data Collection

Emotion Recognition

Evaluation Rating

Object Detection

Prompt Response Writing SFT

Question Answering

Relationship

Segmentation

Text Generation

Transcription

Translation Localization

Freelancer Overview

Born and raised in Hawaii, USA and currently living in Japan. Passionate about building AI that truly understands and elevates humanity, I bring something most annotators simply don’t: deep, authentic empathy combined with obsessive psychological insight. My greatest strength is the ability to connect with people on a profound level — listening not just to words, but to emotional undertones, intent, hesitation, and unspoken meaning. This rare combination allows me to excel at training models on empathy, emotional intelligence, natural conversation flow, behavioral nuance, sentiment depth, ethical alignment, and preference ranking in ways that feel genuinely human rather than mechanical. Whether it’s refining responses for warmth and safety or evaluating how well an AI “gets” a user’s emotional state, I don’t just label data — I infuse it with the human alignment that separates good models from truly transformative ones. Living in Japan for 7 years while learning the language has sharpened my ability to see situations through entirely different cultural lenses, making my annotations richer, more balanced, and less prone to Western-centric bias. With over 600 hours of professional voice-based AI training and annotation using studio-grade equipment (Samson Q9U), I deliver exceptional precision in audio tasks — sharpened by years as a symphonic French horn player and bass choir singer, giving me an ear for tone, rhythm, and emotional delivery that few can match. My multimodal expertise extends to image generation and annotation (selecting best-fit AI portraits, person removal, clothing labeling for Google-style search training) and stems directly from launching and running a successful real estate media company, where I mastered composition, lighting, storytelling, Photoshop, HDR, Premiere Pro, and CapCut. Add entrepreneurial rigor — building SOPs, workflows, hiring systems, and high-stakes team communication — plus a Part 107 drone license and relentless pattern recognition in human behavior and psychology, and you get an annotator who doesn’t just meet quality standards… I consistently raise them. If you’re looking for someone who treats every dataset like a chance to make AI more empathetic, more natural, and more aligned with real people — I’m your candidate.

Labeling Experience

Advanced Audio Quality & Multilingual Annotation

Scale AIAudioSegmentationClassification

Toward the conclusion of the project, I specialized in evaluating multilingual audio clips, leveraging my conversational Japanese proficiency to accurately identify language (e.g., confirming Japanese segments), detect speaker count, estimate distance from microphone, assess background noise levels (including if it was overpowering or distracting), determine audio clarity/quality, and verify correct language identification. I also evaluated segmentation and speaker diarization tasks determining speaker-specific turns, noting overlaps or changes, and ensuring precise boundaries for natural flow. This work required acute listening, pattern recognition (from my music background), and nuanced judgment on human speech realism in varied acoustic environments. My cross-cultural perspective (7+ years in Japan) enabled high-accuracy annotations for diverse, real-world scenarios, contributing to more empathetic, robust, and multilingual speech models.

Toward the conclusion of the project, I specialized in evaluating multilingual audio clips, leveraging my conversational Japanese proficiency to accurately identify language (e.g., confirming Japanese segments), detect speaker count, estimate distance from microphone, assess background noise levels (including if it was overpowering or distracting), determine audio clarity/quality, and verify correct language identification. I also evaluated segmentation and speaker diarization tasks determining speaker-specific turns, noting overlaps or changes, and ensuring precise boundaries for natural flow. This work required acute listening, pattern recognition (from my music background), and nuanced judgment on human speech realism in varied acoustic environments. My cross-cultural perspective (7+ years in Japan) enabled high-accuracy annotations for diverse, real-world scenarios, contributing to more empathetic, robust, and multilingual speech models.

2025

Multimodal Image Labeling and AI-Generated Content Assessment (Including Object Tagging, Quality Evaluation, and Preference Ranking)

Scale AIImageBounding BoxClassification

At the start of the project, I contributed to image-based labeling and assessment tasks, including identifying and labeling specific items/elements (e.g., clothing and objects in Google-style search training), evaluating image quality and realism, comparing generated images for best-fit scenarios (e.g., subject placement in varied environments like parks or pools), and performing person removal/editing decisions. This involved applying strict guidelines to ensure accuracy, spotting subtle details/differences, and providing clear annotations to enhance AI's visual understanding and perception. My background in photography, videography, composition, lighting, and storytelling allowed me to deliver precise, nuanced judgments quickly, earning strong positive feedback that led to reassignment to the specialized voice performance track, where I completed the bulk of my contributions.

At the start of the project, I contributed to image-based labeling and assessment tasks, including identifying and labeling specific items/elements (e.g., clothing and objects in Google-style search training), evaluating image quality and realism, comparing generated images for best-fit scenarios (e.g., subject placement in varied environments like parks or pools), and performing person removal/editing decisions. This involved applying strict guidelines to ensure accuracy, spotting subtle details/differences, and providing clear annotations to enhance AI's visual understanding and perception. My background in photography, videography, composition, lighting, and storytelling allowed me to deliver precise, nuanced judgments quickly, earning strong positive feedback that led to reassignment to the specialized voice performance track, where I completed the bulk of my contributions.

2025

Voice Actor & Script Performer for AI Speech & Conversational Training Projects

Scale AIAudioAudio Recording

Recorded structured conversational audio used to train AI speech and dialogue systems, working session-by-session to deliver clear, natural responses across a wide range of topics. I trained AI with my United States American Accent with professional recording equipment. Followed detailed prompts and performance guidelines while acting out realistic scenarios, maintaining consistency, tone accuracy, and pronunciation quality. The AI models were being trained on natural speaking. Completed extended recording sessions of up to 10 hours per day in a sound controlled environment, ensuring data reliability and adherence to project standards. There were over 30,000 AI Generalists in this project doing other generalist tasks. However, among them were a select few anywhere between 500-800 voice actors at a given time.

Recorded structured conversational audio used to train AI speech and dialogue systems, working session-by-session to deliver clear, natural responses across a wide range of topics. I trained AI with my United States American Accent with professional recording equipment. Followed detailed prompts and performance guidelines while acting out realistic scenarios, maintaining consistency, tone accuracy, and pronunciation quality. The AI models were being trained on natural speaking. Completed extended recording sessions of up to 10 hours per day in a sound controlled environment, ensuring data reliability and adherence to project standards. There were over 30,000 AI Generalists in this project doing other generalist tasks. However, among them were a select few anywhere between 500-800 voice actors at a given time.

2025

Education

Work History